7The cochlear amplifier

TCOuter hair cells that move

In 1985 William Brownell made a startling observation: an isolated outer hair cell changes length when its membrane voltage changes — it shortens when depolarised and lengthens when hyperpolarised. This somatic electromotility is fast, following voltage at acoustic frequencies, and it is unique to outer hair cells. Here was a sensory cell that is also a motor.[1985]

The significance is immediate. Recall from Module 6that the travelling wave deflects the hair bundle and changes the cell's voltage. If that voltage change in turn makes the cell change length, the outer hair cell can push back on the basilar membrane in time with the wave — feeding energy into it rather than just responding to it.

CPrestin, the motor

The molecular motor was identified in 2000: a protein packed in millions of copies into the outer hair cell's lateral wall, named prestin (from presto, fast). Prestin changes its shape directly in response to membrane voltage — a voltage sensor that is also a motor — and the summed conformational changes of all those molecules shorten or lengthen the cell. Unlike muscle, it needs no ATP cycle per contraction, which is how it keeps up with sound.[2000]

How can a protein move on a voltage rather than a fuel? Prestin belongs to the SLC26 family of anion transporters, but it has been repurposed: instead of ferrying chloride and carbonate across the membrane, it holds an intracellular anion as a built-in voltage sensor. When the cell hyperpolarises the anion binds, prestin expands its membrane area and the cell lengthens; when it depolarises the anion leaves, prestin's area shrinks and the cell contracts. Because the trigger is charge displacement, not a chemical cycle, the motor can run at acoustic frequencies — far faster than any ATP-driven cellular motor. Deleting or mutating prestin in mice abolishes electromotility and costs about 40–60 dB of sensitivity.[2010, 2000]

TCThe cochlear amplifier

Put together, this is the cochlear amplifier: at low sound levels the outer hair cells inject mechanical energy into the travelling wave, near its peak, on each cycle. The effect is to amplify the basilar-membrane response by as much as 40–60 dB for soft sounds and to sharpen the tuning of the peak dramatically. The passive mechanics lay out the tonotopic map; the active amplifier makes it sensitive and selective. Explore its input–output behaviour below.[2012]

CPrestin is probably not the only motor. The hair bundle itself can move actively: the same transduction channels and their adaptation machinery can generate force, and in non-mammals — which have no prestin — this active hair-bundle motility is the amplifier outright. In the mammalian cochlea the two mechanisms most likely work side by side, prestin providing the power and the bundle providing fast, frequency-tuned feedback. Either way the principle is the same: the cochlea spends metabolic energy to push back on its own travelling wave.[2010]

CCompression and the dynamic range

Crucially, the gain is level-dependent: large for soft sounds, small for loud ones. This makes the response compressive — a 100 dB range of input is squeezed into a much smaller range of basilar-membrane motion. Compression is how the ear accommodates its vast dynamic range without overwhelming the hair cells and nerve, and it is an active feat: a dead cochlea is linear and insensitive.[2012]

CEfferent gain control

The amplifier is not fixed — the brain turns it down. The medial olivocochlear efferent system projects from the brainstem directly onto the outer hair cells and releases acetylcholine onto an unusual α9/α10 nicotinic receptor found almost nowhere else in the body; the resulting calcium entry activates a potassium current that hyperpolarises the cell and reduces its gain. This olivocochlear reflex is thought to help hearing in noise (by turning down the response to a steady background) and to offer real protection against loud sound — mice lacking the α9 receptor lose that protection and are more easily damaged by noise. It makes the cochlear amplifier a controllable, adaptive gain stage rather than a fixed booster.[2009, 2002]

FTWhen the amplifier fails

Outer hair cells are the most vulnerable cells in the ear — noise, ageing, and ototoxic drugs damage them first. The classic proof came from cats given the ototoxic antibiotic kanamycin, which kills outer hair cells while sparing inner ones: their auditory-nerve fibres lost exactly the low-threshold sensitivity and sharp tuning, while the high-threshold response survived — pinning amplification and fine tuning squarely on the outer hair cells. [1976] Lose them and three things happen together, visible in the widget as you drop their function: thresholds rise (the boost is gone), tuning broadens (frequency selectivity falls), and loudness, once a sound is finally audible, grows abnormally fast — recruitment. This triad is the typical sensorineural hearing loss, and it explains why simply amplifying sound (a hearing aid) only partly helps: it can restore audibility but not the lost compression and sharp tuning.[2012]

FTThe implant has no amplifier

A cochlear implant has no cochlear amplifier — it bypasses the outer hair cells entirely. Electric hearing is therefore notcompressive at the cochlea, and the usable range between a barely-audible and an uncomfortably-loud electrical stimulus is strikingly narrow — often only a handful to a couple of dozen units of current, against the ear's ~120 dB acoustic range. The implant's sound processor must do the cochlea's compression job artificially, mapping the wide acoustic range into that narrow electrical window — the work of front-end compression and dynamic-range settings in programming.[2009]

The amplifier also leaves an audible fingerprint: because it is active and nonlinear, the healthy cochlea actually emits sound. That is the next module.